Cai Meng, Yin Jing, Jin Yi, Liu HongJun
Department of Pain Medicine, Jinling Hospital, Affiliated Hospital of Medical School, Nanjing University, Nanjing, People's Republic of China.
Risk Manag Healthc Policy. 2025 Apr 12;18:1279-1289. doi: 10.2147/RMHP.S503360. eCollection 2025.
Lumbar disc herniation (LDH) usually c auses sciatica. Although selective nerve root block (SNRB) is an effective, highly target-oriented interventional procedure for patients with LDH, accurately predicting the risk of sciatica recurrence in such patients after SNRB remains a major challenge.
We aimed to construct a nomogram model by integrating clinical data, imaging features and inflammation markers that could predict recurrent sciatica following SNRB in LDH patients, which fill the inflammation data gaps during model construction.
In total, 121 sciatica patients were enrolled and assigned to the recurrence group (n = 41) and non-recurrence group (n = 80). By performing the logistic regression analyses, we identified risk factors serving as independent predictors and constructed the nomogram prediction model. Then, the performance and clinical practicality of the nomogram model were validated by performing the receiver operating characteristic curve (ROC) analysis, calibration curve analysis, and decision curve analysis (DCA). The bootstrap method was applied for the internal validation of the nomogram model.
Preoperative sensory symptoms (odds ratio [OR] [95% confidence interval (CI)]: 2.933 [1.211-7.353]), type of herniation (OR [95% CI]: 2.712 [1.261-6.109]), and systemic inflammation response index (OR [95% CI]: 2.447 [1.065-6.271]) were included in the nomogram for predicting unfavorable outcomes following sciatica. The nomogram AUC was 0.764, and the prognostic precision, validated using the bootstrap method, reached 0.756. The ROC and calibration curve analyses, and DCA also produced excellent results, exhibiting favorable predictive performance and clinical benefit.
This study thus identified risk factors that predict unfavorable outcomes after sciatica and developed a risk prediction model based on clinical, radiologic, and inflammatory factors, thereby facilitating early predictions by clinicians and offering an individualized medical interventions for patients with recurrent sciatica in early stages.
腰椎间盘突出症（LDH）通常会导致坐骨神经痛。尽管选择性神经根阻滞（SNRB）是一种针对LDH患者有效的、高度靶向性的介入治疗方法，但准确预测此类患者SNRB后坐骨神经痛复发的风险仍然是一项重大挑战。
我们旨在通过整合临床数据、影像特征和炎症标志物构建一个列线图模型，以预测LDH患者SNRB后复发性坐骨神经痛，填补模型构建过程中的炎症数据空白。
共纳入121例坐骨神经痛患者，分为复发组（n = 41）和非复发组（n = 80）。通过进行逻辑回归分析，我们确定了作为独立预测因素的危险因素，并构建了列线图预测模型。然后，通过进行受试者工作特征曲线（ROC）分析、校准曲线分析和决策曲线分析（DCA），验证了列线图模型的性能和临床实用性。采用自助法对列线图模型进行内部验证。
术前感觉症状（比值比[OR][95%置信区间（CI）]：2.933[1.211 - 7.353]）、突出类型（OR[95%CI]：2.712[1.261 - 6.109]）和全身炎症反应指数（OR[95%CI]：2.447[1.065 - 6.271]）被纳入列线图，用于预测坐骨神经痛后的不良结局。列线图曲线下面积（AUC）为0.764，采用自助法验证的预后精度达到0.756。ROC和校准曲线分析以及DCA也产生了优异的结果，显示出良好的预测性能和临床效益。
本研究因此确定了预测坐骨神经痛后不良结局的危险因素，并基于临床、放射学和炎症因素开发了一个风险预测模型，从而便于临床医生进行早期预测，并为早期复发性坐骨神经痛患者提供个体化医疗干预。
